Why are all tweeters 1 inch in diameter ?

[Pearljam5000]

Sorry for the dumb question but i always wondered why all tweeters are 1 inch?
Let's say they will make it 2-3 inches, will it sound better?

 

[sergeauckland]

It's for several reasons.
Firstly, they're small to keep the mass low, as high mass means low sensitivity
Secondly, dispersion. A small source has wider dispersion than a large source due to path length differences.
Thirdly, stiffness as it's easier to make something small stiff than a larger area, thus any resonances and breakup can hopefully be above the audible range.

On the other hand, if too small, then they need more excursion to move enough air to be loud enough, so somewhere between 3/4"-1 1/2" is about right.

S

 

[Rock Rabbit]

As any kind of loudspeaker driver they radiate efficiently as a "piston" when driver perimeter is comparable to a wavelength. Then 1" is good over 4 kHz and acceptable with little error at one lower octave or 2 kHz covering the typical range of a tweeter.
They don't require a big effective cone area because of better efficiency at high frequency (acoustic resistance proportional to f^2) and directionality effects with wavelength small than half perimeter or 8 kHz.

 

[RealityCheck18]

A few additional factors,

Beaming--a 1" dome beams at over 14KHz which is good enough (look at dispersion charts for results) You can get dome tweeters from 10mm to 30mm but the tiny 10mm versions have very small voice coils so can't handle much power. The larger 30mm ones are generally used for lower crossover points of around 2,000 Hz but will beam earlier--physics can't be denied! The 1" or 25mm dome is a good compromise between efficiency, power handling beaming etc. so quite common.

"All" tweeters are not 1", but most consumer dome speaker tweeters are. Once you jump into compression drivers, the most common is a 1" throat with a compression driver "dome" size of between 1.5 to 2" (38 to 50mm) They use phase plugs etc. at usually make it to around 18KHz without too much trouble. Most of those 1" throat drivers cross at 1,600Hz or higher but their are exceptions, some of them can be crossed over at lower frequencies although you reduce their power capacity--common in consumer speakers that use those drivers. For example, the B&C DE250 normally crosses over at 1,600 Hz at full power (60 watts) but some designs for consumer/home theater use drop it down to 1,200Hz, 1,000 Hz or even--when used with huge horns/waveguides--down to 850 Hz (Earl Geddes does this)

Once you go past around 30mm for a conventional radiating dome, the beaming and respose gets poor over 10,000Hz so most of the 2" domes you see are classified as midranges. For compression drivers, you can increase the "throat" size to 1.4" or 2" but that increases beaming so the large 3" or 4" "domes" in those massive compression drivers are mated to huge horns and used as midranges.

It would be cool if a 3" dome didn't beam--but they do! Infinity back in the 80's made 5" domes and 3" domes for their Kappa series of speakers. The 5" domes were "lower mids" and the 3 inchers were "upper mids" in their 5 way Kappa 8 and Kappa 9 speakers. Big domes = Big hair so now you know! There is a method to the madness with driver design, once you go against the brick wall of physics--then make your compromises accordingly to whatever design you like.

Hope that helps

 

[AnalogSteph]

Most of those 1" throat drivers cross at 1,600Hz or higher but their are exceptions, some of them can be crossed over at lower frequencies although you reduce their power capacity--common in consumer speakers that use those drivers. For example, the B&C DE250 normally crosses over at 1,600 Hz at full power (60 watts) but some designs for consumer/home theater use drop it down to 1,200Hz, 1,000 Hz or even--when used with huge horns/waveguides--down to 850 Hz (Earl Geddes does this)

This variability is coming about because the horn / waveguide is providing impedance matching between the driver and surrounding air, acoustic impedance being the ratio of pressure and velocity. A compression driver does not have too much surface area or excursion, but it can exert considerable pressure. This means a characteristic impedance that's way different from free air, which you may have noticed does not take very much pressure to move quite a bit. A waveguide transforms impedance to be more suitable to transmission, effectively increasing sensitivity. A larger one remains effective to lower frequencies, hence reducing excursion. This is most easily understood if you have an idea of how RF propagation works.

You can think of impedance matching as what the transmission in a vehicle does (well, at least for any given frequency). It translates an RPM of best long-term engine power output into optimum speed while countering whatever opposing forces the vehicle is subjected to. (Cue flashbacks to the relationship between force, energy and power.)
If you need to drag a tractor up a steep hill, very short gearing is needed since a substantial part of gravitational force is opposing movement and a considerable amount of potential energy has to be pumped into the system in a given time even when moving relatively slowly.
Our speaker example is much more like driving on a perfectly level highway though, where you are only fighting friction losses in wheels and bearings (approximately linear with speed) and aerodynamic drag (approximately proportional to speed squared), forces that are only reaching the same magnitude at much higher speeds. This culminates in some sports coupés with large displacement motors (>5 liters) essentially being able to sustain US highway speeds at idle or little more in tallest gear, with actually rather decent fuel mileage indeed. That's basically what a compression driver in a horn is, with fuel mileage being the equivalent of sensitivity.

We aren't using compression drivers with horns / waveguides everywhere because at some point they become uncomfortably large, and you tend to be overall better off with cone drivers of substantial surface area and excursion which are providing a lot of displacement right out of the box. (Note that even these can be used with extra impedance matching if so desired, and in fact this is being used in large high sensitivity speakers.)

 

[Vini darko]

I have 10cm planar tweeters. It's all beaming

 

[RayDunzl]

Sorry for the dumb question but i always wondered why all tweeters are 1 inch?

My "tweeters" cover 15 x 48 inches.

Ok, they are really about 4 x 15 but twelve are stacked vertically.

Let's say they will make it 2-3 inches, will it sound better?

I suspect they wouldn't go very loud here.

---

Sorry for the dumb answers

 

[amirm]

Good topic and answers. Promoted to home/review page.

 

[beagleman]

I have owned tweeters in vintage speakers that were 3" and 5" in diameter. (Different speakers)

They sounded "okay" but the dispersion fell off at high frequencies quite a bit.
They were not as airy or great at transients either.

 

[DonH56]

The usual answer about beaming (reduced dispersion angle) is that it starts when the frequency is such that the tweeter's diameter (or axis width for planar and ribbon tweeters) is one wavelength. Using 1127 ft/s or 13,524 inch/s for the velocity of sound, a 1" diameter piston yields 13,524 Hz. The actual diameter of the moving piston is probably a little less, of course, so the frequency will be a little higher. This works for woofers and midrange drivers as well.

Metric folk can use 34,351 cm/s for the speed of sound.

 

[AudioSceptic]

The usual answer about beaming (reduced dispersion angle) is that it starts when the frequency is such that the tweeter's diameter (or axis width for planar and ribbon tweeters) is one wavelength. Using 1127 ft/s or 13,524 inch/s for the velocity of sound, a 1" diameter piston yields 13,524 Hz. The actual diameter of the moving piston is probably a little less, of course, so the frequency will be a little higher. This works for woofers and midrange drivers as well.

Metric folk can use 34,351 cm/s for the speed of sound.

I love the precision of the numbers, but I doubt you can be that precise without knowing air temperature and pressure. ;-)

 

[DonH56]

I love the precision of the numbers, but I doubt you can be that precise without knowing air temperature and pressure. ;-)

Of course, but how detailed did you want? This was sea level, dry air, standard temp and pressure. Humidity is also a first-order effect. Feel free to provide the whole equation; I was too lazy to look it up in my old acoustics text so just spouted off what I remembered off-the-cuff. And of course you don't go from wide dispersion to beaming immediately, there's a transition period. And so forth. I just wanted to put some ballpark numbers out there for people to calculate on their own. The only one I remember is 1127'/s; the rest of the digits are from the calculator.

 

[BYRTT]

Sorry for the dumb question but i always wondered why all tweeters are 1 inch?
Let's say they will make it 2-3 inches, will it sound better?

Modeled in CAD software i get these off axis curves for 19mm / 25mm / 50mm / 75mm pistons, as seen 2-3 inch diameters are not a very modern performance ..

 

[Newman]

The usual answer about beaming...is that it starts when the frequency is such that the tweeter's diameter (or axis width for planar and ribbon tweeters) is one wavelength.
.... a 1" diameter piston yields 13,524 Hz.
... This works for woofers and midrange drivers as well.

That’s how I see it too. It’s a rule of thumb but a handy one.

A 1” / 25mm tweeter ‘starts to beam’ at 13-14 kHz, and our hearing doesn’t seem to care that much if moderate beaming is occurring from 15-20 kHz.

For perfectionists there is the 3/4” / 19mm tweeter that ‘starts to beam’ at 18-19 kHz. But it generally needs a higher crossover frequency Iike 3-5 kHz, so you can’t use it in a 2-way speaker. You really need 3-way speakers if you want a tweeter smaller than 1” / 25mm. That’s the real reason why 1” / 25mm tweeters are the default size.

If an electrostatic panel 400mm wide is left unsegmented, it ‘starts to beam’ at 850 Hz, and by the time you get to treble frequencies it is basically a LASER beam, and a bit of a joke. That is why panel speakers tend to be segmented within the one panel, or use a separate narrow panel for treble, with a crossover. The Quad ESL used circular segments and the treble segment was 4” / 100mm diameter, so would ‘start to beam’ at 3-4 kHz, which is still unsatisfactory.

cheers

 

[Alice of Old Vincennes]

My "tweeters" cover 15 x 48 inches.

Ok, they are really about 4 x 15 but twelve are stacked vertically.

View attachment 106544




I suspect they wouldn't go very loud here.

---

Sorry for the dumb answers

They will play loud enough with Crown XLS 2502 and will suck up every watt. I attached my neck to a cervical collar and braced with 2 x 4's to stay in the sweet spot. I left the Maggie experiment. My wife and kids could not fit in that special spot.

 

[Ron Texas]

How big is a horn tweeter? Is it the driver or the exit from the horn?

 

[escape2]

Let's say they will make it 2-3 inches, will it sound better?

Flat panel / ribbon tweeters are usually about that size. Do they sound better? I'd say it depends on the specific one.

My Chane A2.4 has a 1.25" x 2.25" flat panel leaf tweeter. It sounds great to me.

 

[typericey]

What boggles my mind more is how a 1" driver can play so loud. In some configurations it can run alongside a couple of midranges and a few woofers in a large floorstander with upwards of 250W power handling rating. How can annoyingly loud cymbal crashes come cleanly out of a 1 inch thing with <1mm of excursion?

 

[Duke]

How big is a horn tweeter? Is it the driver or the exit from the horn?

If you're asking about the radiation pattern, that is largely set by the shape of the horn. To a reasonable first approximation, the coverage pattern of a horn is predicted by the angle of the walls of the horn.

In practice the coverage pattern changes with frequency for some types of horns (typically narrowing as we go up in frequency), and for others remaining fairly uniform across most of the spectrum that the horn covers. The latter are often referred to as "constant directivity" horns, and the horn in the JBL M2 would be an example.

I'm leaving a lot out for the sake of brevity.

 

[BYRTT]

...How can annoyingly loud cymbal crashes come cleanly out of a 1 inch thing with <1mm of excursion?

Probably the crash itself is lower in frequency and routed/filtered to midwoofer plus midrange so tweeter only handles the splashes but not the crashes ..